Course Introduction

Module 0

Prof. Matthew G. Son

University of South Florida

Course Description

About this course

  • Learn the fundamental technology in FinTech

    • The language of data scientists
    • Concepts of Big data in Finance
    • Explore various ML algorithms for Finance

About this course

A “Hands on” course

  • Lab sessions and exercises in the class

  • Emphasize on building practical skills (and good habits)

    • Process and Summarize Financial Data

    • Practical knowledge in how ML algorithms work

About me

Matthew Son, Ph.D. in Finance

  • Office: BSN 3127

  • gson@usf.edu

  • R, Python, C/C++

  • Research area:

    • Machine learning / Big data in Finance
    • Market Microstructure, Asset Pricing
    • Derivatives, Fixed Income

About you

Please tell us about yourself briefly:

  • Your name, major
  • Your career experiences if any, background
  • Your programming skills
  • Your interest / goal

Course Goal

The goal of this course is:

  • To equip technical proficiency to work on financial data
  • Knowledge and basic skills for big data processing
  • Understand how ML algorithms work

and ultimately:

  • Ablility to make a financial report using ML

Technology

  • Throughout the course you’ll use and learn:

    • R programming language and packages

    • VScode IDE for main interface

    • Quarto for technical documentation

    • h2o.ai for ML algorithm implementations

    • Copilot coding agent tools for programming

Prerequisites

  • Basic knowledge in Investments (Asset pricing) and Corporate Finance

  • Excel & Financial calculator

  • General proficiency with computers

  • Some experience in any programming

    • Self-guided study is strongly recommended

Structure of the class

  • A two-part structure: lectures and lab sessions

  • Lecture: the instructor will cover the concepts and demonstrate the workflow in R.

    • Students are encouraged to engage by copying, typing, and running code on their own.

  • Lab sessions:

    • Students will work on coding problems that are closely tied to the lecture topics.

Coding questions

In person is preferred over email.

  • TA also can provide help you in-person and virtually

When writing in email, include:

  • What is the error message (the response)
  • What you have tried
  • Prepare reproducible example (Reprex)

Course Materials

Textbook

No required textbook. Lecture notes and supplementary materials will be provided throughout the course.

1. John C. Hull, “Machine Learning in Business: An Introduction to the World of Data Science”, 3rd edition, 2021, GFS Press. ISBN-13: “979-8508489441”

2. Darren Cook, “Practical Machine Learning with H2O”, 1st edition, 2017, OReilly. ISBN-13: “978-1491964606”

3. “Python Polars: The Definitive Guide”, 1st edition, 2023, OReilly. ISBN-13: “978-1098156084”

4. Hadley Wickham & Garrett Grolemund, R for Data Science, 2nd edition, 2023, OReilly. ISBN-13: “978-1491910399”. Electronic copies are available for free.

Computer & Software

  • The latest stable version of R and VScode.

  • Please bring your laptop (macOS/Windows/Linux with popular distro)

Grading

Grading Categories and Weights

Graded Items Percent of Final Grade
Participation 10%
ML Assignment 15%
In-Class Quizzes 30%
Lab Problems 15%
Final Exam 30%

Course Outlines

Course Modules

  • Module 0: The FinTech Landscape

  • Module 1: Financial analysis with R

  • Module 2: Introduction to Big data analytics

  • Module 3: Unsupervised Learning

  • Module 4: Supervised Learning

Course Schedule

The course schedule is tentative and subject to change.

Week Topics Finance Applications
1 ~ 4 R & Quarto

  • Financial data manipulation basics

  • Technical documentation for Finance

  • Pivot & Tidying Financial data

  • Visualizations

  • Time-series visualizations

  • CAPM

  • Portfolio performance analysis
5 ~ 7 Big data
  • Financial database / Big data workflows
  • SQL / Database backends
8 ~ 10 Unsupervised Learning

  • Introduction to Machine Learning in Finance

  • Unsupervised Learning: K-means clustering, PCA

  • Credit risk analysis with ML

  • Stock classifications with ML
11 ~ 15 Supervised Learning
  • Supervised Learning
  • Linear Regression & logit regression
  • Decision trees
  • Ensemble Methods
  • Random Forests / Gradient Boosting
  • Fama-French Factor Models
  • Real Estate Pricing
  • Credit decision models

Course Policy

Attendance

Attendance is expected.

  • Random attendance check for bonus points
  • Active students will receive bonus points

Quizzes

In-class quizzes will be administered through Canvas with Honorlock monitoring.

  • Basic conceptual multiple-choice questions
  • True/False questions, matching questions
  • Simple coding questions (emphasis on reading)

Closed book, 1 page of cheat sheet (can be typed) is allowed.

ML Assignment

  • You will be provided a financial dataset.

  • Explore, visualize and provide insights about the data

  • Apply machine learning to develop a model and evaluate

  • Document should be generated using Quarto.

  • No late submissions will be accepted.

Final Exam

  • In persaon Canvas Exam and Honorlock screen monitoring

  • No AI tools are allowed

  • Closed book

  • A letter-size, double-sided cheat sheet will be allowed.

  • 2 hours

During Class

  • Ensure not to disrupt the course by talking, arriving late, eating, etc.

  • Please limit computer usage to activities directly related to the class.

  • Phones are not permitted as they are unlikely to be useful for course-related activities.

Late submissions

I won’t grade late submissions except:

-   if a valid excuse is communicated to the instructor before the deadline

-   valid excuses with proof will be accepted later, in extenuating circumstances

Missing exam, quiz

A valid excuse must be communicated to the instructor before the exam/quiz

  • In such cases, a make-up exam/quiz may be scheduled prior to the original exam/quiz date

Request for Re-evaluation

Students may request re-grading exams and assignments within one week (seven calendar days) after grading.

In the case of a regrading request after the final exam, all previous submissions for the course will be thoroughly reevaluated to ensure consistent grading standards.

For more details: